Electrocatalytic deposition of silver onto AuNPs applied to magneto immunoassays

In this work, an electrocatalytic silver-enhanced metalloimmunoassay using AuNPs as labels and microparamagnetic beads (MB) as platforms for the immunological interaction is developed for model proteins, in order to achieve very low detection limits with interest for further applications in several fields.

3.2.1. Catalytic effect of AuNPs on the silver electro-deposition.

The silver enhancement method, based on the catalytic effect of AuNPs on the chemical reduction of silver ions, has been widely used to improve the detection limits of several metalloimmunoassays. In these assays (Karin et al., 2006; Guo et al., 2005; Chu et al., 2005) the silver ions are chemically reduced onto the electrode surface in the presence of AuNPs connected to the studied bioconjugates, without the possibility to discriminate between AuNP or electrode surface. Furthermore, these methods are time consuming and two different mediums are needed in order to obtain the analytical signal:

the silver/chemical reduction medium to ensure the silver deposition and the electrolytic medium necessary to the silver-stripping step.

However, in this work, for the first time, the selective electro-catalytic reduction of silver ions on AuNPs is clarified, and the advantages of using MBs as bioreaction platforms combined with the electrocatalytic method are used to design a novel sensing device.

The principle of the electrocatalytic method is resumed in figure 1A. Cyclic voltammograms, obtained by scanning from +0.30 V to –1.20 V in aqueous 1.0 M NH3 / 2.0 x 10^-4 M AgNO3, for an electrode without (a) and with (b) AuNPs previously adsorbed during 15 minutes are shown. It can be observed that the half-wave potential of the silver reduction process is lowered when AuNPs are previously deposited on the electrode surface. Under these conditions, there is a difference (ΔE) of 200 mV between the half-wave potential of the silver reduction process on the electrode surface without (a) and with (b) AuNPs are adsorbed on the electrode surface (b). The amount of the catalytic current related to silver reduction increases with the amount of AuNPs adsorbed on the electrode surface (results not shown).

M8

Fig. 1. (A) Cyclic voltammograms, scanned from +0.30 V to –1.20 V in aqueous 1.0 M NH3-2.0 x 10^-4 M AgNO3, for an electrode without deposited AuNPs (a-curve) and for an potential: -0.12 V; silver deposition time: 60 seconds; scan rate: 50 mV/s.

Taking this fundamental behavior into account, a novel analytical procedure for the sensitive detection of AuNPs is designed. It consists in choosing an adequate deposition potential, i.e. -0.12 V, at which the direct electro-reduction of silver ions, during a determined time, would take place on the AuNPs surface instead of the bare electrode surface. At the beginning of the process, the electrocatalytical reduction of silver ions onto the AuNPs surface occurs and once a silver layer is already formed more silver ions are going to be reduced due to a self-enhancement deposition. The electrocatalytic process is effective due to the large surface area of AuNPs allowing an easy

-10

! 43!

diffusion and reduction of the silver ions. The proposed mechanism is the following:

In a first step, while applying a potential of -0.12 V during 60 s, the silver from the ammonia complex, are reduced to a metallic silver layer onto the AuNPs surface:

AuNPs, E= -0.12 V

Ag(NH3)2+ + 1 e^- → Ag(deposited on AuNPs) + 2 NH3

In a second step an anodic potential scan is performed (from -0.12 V to +0.30 V) in the same medium, during which the re-oxidation of silver at +0.10 V is recorded:

E(-0.12 V to +0.30 V) Ag(deposited onto AuNPs) + 2 NH3 → Ag(NH3)2 + 1 e

-The amount of silver electrodeposited at the controlled potential (corresponding to deposition onto AuNPs surface only) is proportional to the adsorbed AuNPs. Consequently the re-oxidation peak at +0.10 V produces a current which is proportional to the AuNPs quantity. The obtained re-oxidation peak constitutes thus the analytical signal, used later on for the AuNPs and consequently the protein quantification.

3.2.2. Sandwich type immunocomplex

The preparation of the sandwich type immunocomplex was carried out following a previously optimized procedure (Ambrosi et al., 2007), but introducing slight changes in order to minimize the unspecific absorptions that interfere the sensitive electrocatalytic detection. The analytical procedure is schemed in figure 2 (for detailed experimental conditions, see Publication 5 in Chapter 7).

The use of blocking agents so that any portion of the MB surface which does not contain the primary antibody is "blocked" thereby preventing non-specific binding with the analyte of interest (protein) is crucial. The obtained values of the analytical signals are highly dependent on the blocking quality. Following the previously reported procedure, based on the direct electrochemical detection of AuNP, the blank samples signal (the samples without the antigen or with a non-specific antigen) were very high (data not shown).

! 44!

Fig. 2. Schematic (not in scale) of: (A) AuNP conjugation with anti-human IgG; (B) Analytical procedure for the sandwich type assay and the obtaining of the analytical signal based on the catalytic effect of AuNPs on the silver electrodeposition. Procedure detailed in experimental section.

The resulted unspecific adsorptions could be due to some factors. For a given concentration of the blocking agent the unspecific adsorptions will depend on the time interval used to perform such a step. By increasing the time interval of the blocking step (from 30 to 60 minutes and using PBS-BSA 5% as blocking agent) in the sandwich assay we could ensure a better coverage of the free bounding sites onto the MB surface avoiding by this way the unspecific adsorptions. Another important factor that affects the unspecific adsorptions is the washing step that aims at removing the unbound species avoiding by this way possible signals coming from AuNPs not related to the required antigen. Stirring instead of gentle washing brought significant decrease of unspecific adsorptions too. TEM images of the sandwich assay before and after the mentioned improvement corroborated also in understanding the phenomena related to these non-desired adsorptions (see Publication 4, Supplementary Info. to more information).

Clear evidences of the successful immunological reaction in a condition of the absence of unspecific adsorptions are the transmission electron micrographs (TEM) images shown in figure 3.

! 45!

Fig. 3. Transmission electron micrographs (TEM) images of the MBs after the sandwich type assay detailed in experimental section. Assay carried out with 1.0 x 10-3 μg/ml of the non specific antigen (goat IgG) (blank assay-A-) and assay performed with 1.0 x 10-3 μg/ml of the specific antigen (human IgG) (B).

When the assay is carried out in the presence of the non-specific antigen (goat IgG - Figure 2A) only MBs are observed with a very low amount of AuNPs non specifically bonded. However, if the assay is performed with the specific antigen (human IgG - Figure 2B), a high quantity of AuNPs is observed around the MBs, which indicates that the immunological reaction has taken place.

Thus, following the analytical procedure described in experimental section the selective silver deposition onto the AuNPs surface is achieved. The potential and the time of the silver electro-deposition have been previously optimized (see figure S2 in the supplementary material). The application of a -0.12 V potential during 60 s resulted the best as a compromise between the higher sensitivity and analysis time.

Typical analytical signals obtained for the sandwich type assay performed with a human IgG concentration of 5.0 x 10^-7 μg/mL (a) and for the blank assay performed with the same concentration of goat IgG (b) are shown at figure 3B.

The electrocatalytic deposition of silver ions onto the surface of the magnetic electrode versus the applied potential used for silver deposition is studied also by scanning electron microscopy (SEM). Figure 4 shows SEM images of the MB deposited onto the magnetic electrode surface, after performing silver

M/

electro-deposition at different deposition potentials (-0.10, -0.12 and -0.20 V) during 1 minute.

Fig. 4. Scanning electron microscopy (SEM) images of the MB deposited on the electrode surface, after the silver electro-deposition in aqueous 1.0 M NH3-2.0x10-4 M AgNO3, at -0.10 V (A,B,) -0.12 (C,D) and -0.20 V (E,F) during 1 minute, for the sandwich type assay performed as described in experimental section, with the non specific antigen (goat IgG -blank assays-A,C,E) and for the specific antigen (human IgG) at concentration of 1.0 x 10-3 μg/mL (B,D,F).

The upper images of Figure 4 (A,C,E) correspond to the sandwich type assays performed with the non specific antigen (goat IgG - blank assays). The lower part images (B,D,F) correspond to the assays with an specific antigen (human IgG) concentration of 1.0·10-3 μg/mL. It can be observed that at a deposition potential of -0.10 V, (A) no silver crystals are formed in the absence of the specific antigen while low amounts of silver crystals (white structures in the B image) are observed with the assay performed with the specific antigen. This means that the silver deposition has scarcely occurred to the AuNPs anchored onto the MB through the immunological reaction. (B).

The formation of these silver crystals is much more evident when the same assay (with specific antigen) is performed at deposition potential of -0.12 V (D) where a bigger amount of MB appear to be covered with silver crystals – the same phenomena not observed for the blank assay (C). The obtained image is clear evidence that the used potential have been adequate for the silver deposition onto the AuNPs attached to the MB through the immunological reaction. The Energy Dispersive X-Ray (EDX) analysis (provided by SEM instrument) is also performed and the results are in

MH negative deposition potentials (i.e. -0.20 V) the deposition of silver takes place in a high extent also on the electrode surface as it was expected (E). This phenomenon can be appreciated by bigger cluster like white silver crystals that may be associated not only with silver deposited onto the AuNPs but also onto the surface of the magnetic electrode. This potential (-0.20 V) is not adequate to quantify the specific antigen due to false positive results that can be generated. The -0.12 V have been used in our experiments as the optimal deposition potential that can not even discriminate between the assays and the blank but also be able to do protein quantification at a very low detection limit.

Similar silver structures formed onto AuNPs have been earlier after chemical silver (I) reduction for a DNA array-based assay (Park et al., 2002) or an immunoassay (Gupta et al., 2007) but this is the first time that such potential controlled silver deposition induced by the electrocatalytical effect of AuNP are being evidenced. Moreover the relation between the current produced by the oxidation of the selectively deposited silver layer and the quantity of AuNP is demonstrated as see in the following part.

Fig. 5. (A) Cyclic voltammograms recorded in aqueous 1.0 M NH3-2.0 x 10-4 M AgNO3, from -0.12 V to +0.30 V, for the sandwich type assay described in experimental section with 1.0 x 10-6 μg/mL of the non specific antigen (goat IgG -thin line) and for increasing specific antigen (human IgG) concentrations: 5.0 x 10-8, 1.0 x 10-7, 5.0 x 10-7, 7.5 x 10-7 and 1.0 x 10-6 μg/mL. Silver electro-deposition potential: -0.12 V; silver deposition time: 60 seconds; scan rate: 50 mV/s.

In figure 5A are shown cyclic voltammograms for different concentrations of human IgG following the procedure explained in experimental section. Figure 5B represents the corresponding peak heights used as analytical signals,. As observed in this figure a good linear relationship for the concentrations of

-10 -5 0 5 10 15 20 25 30

-0.15 -0.05 0.05 0.15 0.25 0.35

Potential / V Current / µA

Ms

human IgG, in the range from 5.0 x 10-8 to 7.5 x 10-7 μg/mL, with a correlation coefficient of 0.9969, according to the following equation:

ip (μA) = 21.436 [human IgG] (μg/mL) + 3.750 (n = 3) is obtained.

Fig. 5 (B) The corresponding relationship between the different concentrations of the human IgG and the obtained peak currents used as analytical signals.

The limit of detection (calculated as the concentration corresponding to three times the standard deviation of the estimate) of the antigen was 23 fg ofhuman IgG for mL of sample. The reproducibility of the method shows a RSD around 4%, obtained for a series of 3 repetitive immunoreactions for 5.0 x 10-7 μg human IgG / mL.

These results indicate that with the silver enhancement method can be detected 1000 times lower concentrations of antigen than with the direct differential pulse voltammetry (DPV) gold detection as done previously in our group.

0 5 10 15 20 25

0.00 0.20 0.40 0.60 0.80 1.00 1.20

[Human IgG] / ng mL^-1 ip /µA

! 49!

3.3. Electrochemical quantification of AuNPs based on the